The present invention is concerned with a novel intermediate and process for synthesizing compounds which inhibit the protease encoded by human immunodeficiency virus (HIV), and in particular L-735,524, or pharmaceutically acceptable salts thereof. These compounds are of value in the prevention of infection by HIV, the treatment of infection by HIV and the treatment of the resulting acquired immune deficiency syndrome (AIDS).
More specifically, the instant process involves the reaction of an amine nucleophile such as a piperazine derivative, with an activated glycidyl derivative such as 2(S)-glycidyl nosylate, to afford an epoxide intermediate which is useful in the preparation of HIV protease inhibitor compounds, including L-735,524. Also provided is an improved process for the synthesis of specific dialkylamines used in the synthesis of HIV protease inhibitors.
A retrovirus designated human immunodeficiency virus (HIV) is the etiological agent of the complex disease that includes progressive destruction of the immune system (acquired immune deficiency syndrome; AIDS) and degeneration of the central and peripheral nervous system. This virus was previously known as LAV, HTLV-III, or ARV. A common feature of retrovirus replication is the extensive post-translational processing of precursor polyproteins by a virally encoded protease to generate mature viral proteins required for virus assembly and function. Inhibition of this processing prevents the production of normally infectious virus. For example, Kohl, N. E. et al., Proc. Nat'l Acad. Sci. 85, 4686 (1988) demonstrated that genetic inactivation of the HIV encoded protease resulted in the production of immature, non-infectious virus particles. These results indicate that inhibition of the HIV protease represents a viable method for the treatment of AIDS and the prevention or treatment of infection by HIV.
The nucleotide sequence of HIV shows the presence of a pol gene in one open reading frame Ratner, L. et al., Nature, 313, 277(1985)!. Amino acid sequence homology provides evidence that the pol sequence encodes reverse transcriptase, an endonuclease and an HIV protease Toh, H. et al., EMBO J. 4, 1267 (1985); Power, M. D. et al., Science, 231, 1567 (1986); Pearl, L. H. et al., Nature 329, 351 (1987)!. The end product compounds, including L-735,524 which is shown in Example 11 below, that can be made from the novel intermediates and processes of this invention are inhibitors of HIV protease, and are disclosed in EPO 541,168, which published on May 12, 1993.
Previously, the synthesis of L-735,524 and related compounds was accomplished via a 12-step procedure which employed a hydroxy protected dihydro-5(S)-hydroxymethyl-3(2H)-furanone which was alkylated, and involved replacement of an alcohol leaving group on the alkylated furanone with a piperidine moiety. The coupled product was then hydrolyzed to open the furanone ring into a hydroxy acid moiety, and the acid was ultimately coupled to 2(R)-hydroxy-1(S)-aminoindane. This procedure is described in EPO 541,168. The extreme length of this route (12 steps), renders this process time consuming and labor intensive, and it requires the use of many expensive reagents and an expensive starting material. A route requiring fewer reaction steps and reagents would provide desirable economical and time-saving benefits.
A modified route to L-735,524 and related compounds was also shown in EPO 541,168 based on the diastereoselective alkylation of the enolate derived from N-(2(R)-hydroxy-1(S)-indan-N,O-isopropylidene-yl)-3-phenyl-propaneamide, in which the C.sub.3 -C.sub.5 three-carbon unit was introduced as an allyl group and later oxidized. Some problems with this route are: (a) four steps are necessary to effect the introduction of the three carbon glycidyl fragment, (b) highly toxic OsO.sub.4 is used in the process and (c) low diastereoselectivity is obtained in the dihydroxylation step. Thus, a desirable process would directly introduce the three carbon unit in the correct chiral oxidized form.
Furthermore, the synthesis of the chiral piperazine intermediate was effected from 2-pyrazinecarboxylic acid in a 6 step procedure and required the use of expensive reagents such as BOC-ON and EDC. A shorter route to the piperazine intermediate which also does not use expensive reagents would thus be desired.
Several examples of condensations of stabilized carbanions with glycidol and its derivatives (activated or unactivated) are known in the literature; however, no known methods produce a new epoxide directly in good yield. See, e.g., Hanson, R. M., Chem. Rev., 1991, 91, 437-475. In the case of activated glycidol derivatives and carbon nucleophiles, this is due primarily to the anticipated and undesirable "double" addition of the nucleophile to the epoxide product.
Condensations of stabilized carbanions with activated non-racemic glycidol derivatives have been demonstrated: malonate anion has been coupled with both non-racemic epichlorohydrin i and non-racemic glycidyl triflate ii to afford the cyclopropyl-lactone iii. See, e.g., Pirrung, M. C., et al., Helvetica Chimica Acta 1989, 72, 1301-1310, and Burgess, K., et al., J. Org. Chem. 1992, 57, 5931-5936. Thus, in this case, the intermediate epoxide undergoes further reaction to afford the cyclopropyl ring system. In the case of i, the initial reaction with malonate anion occurs at the epoxide terminus (C.sub.3), whereas with ii, the initial reaction occurs at the triflate C.sub.1 terminus. ##STR2##
A related example is reaction of sulfone-stabilized carbanion derived from v with glycidyl tosylate iv to afford the hydroxy-tosylate vi. See, e.g., Baldwin, J. E., et al., J. Chem. Soc. Chem. Commun. 1992, 1249-1251. In this case, although double addition of the carbanion is not a major problem, an additional step is necessary to convert the intermediate hydroxy-tosylate vi to the desired epoxide vii. ##STR3##
Similarly, it is unknown in the literature and unexpected that nitrogen nucleophiles can be selectively added to activated glycidol derivatives in good yield without the problematic double addition.
Also known in the art is the condensation of amide enolates derived from N-(2(R)-hydroxy-1(S)-indan-N,O-isopropylidene-yl)-3-phenylpropaneamide 7 with protected alpha-amino epoxides viii to afford the desired hydroxyethylene dipeptide isostere intermediates ix with a high degree of stereocontrol for the C.sub.2 -(R)-stereocenter. See, e.g., Askin, D., et al., J. Org. Chem., 1992, 57, 2771-2773 and U.S. Pat. No. 5,169,952 to Askin, D., et al. After hydrolysis, the deprotected hydroxyethylene dipeptide isostere inhibitors are obtained. ##STR4##
The resolution of 2-piperazinecarboxylic acid with (+)-CSA is known. See. e.g., Felder, E., etal., Helvetica Chim. Acta, 1960, 43, 888. However, examples of the resolution of piperazine amides are not known in the literature.
The instant invention provides a more advantageous method for preparing HIV protease inhibitors than previously known. It allows a much shorter, more highly diastereoselective, higher yielding synthesis of the compounds disclosed in EPO 541,168, and in particular L-735,524, without the use of toxic reagents such as osmium tetraoxide or expensive reagents such as (S)-(+)-dihydro-5-(hydroxy-methyl)-2(3H)-furanone.